Vol. 176, No. 11JOURNAL OF BACrERIOLOGY, June 1994, p. 3111-31160021-9193/94/$04.00+ 0Copyright ©D 1994, American Society for Microbiology

Growth and Bioenergetics of Alkaliphilic Bacillus finmus OF4in Continuous Culture at High pH

MICHAEL G. STURR, ARTHUR A. GUFFANTI, AND TERRY A. KRULWICH*Department of Biochemistry, Mount Sinai School of Medicine of the City University

ofNew York, New York, New York 10029

Received 24 January 1994/Accepted 17 March 1994

The effect of external pH on growth of alkaliphilic Bacillus firmus OF4 was studied in steady-state,pH-controlled cultures at various pH values. Generation times of 54 and 38 min were observed at external pHvalues of 7.5 and 10.6, respectively. At more alkaline pH values, generation times increased, reaching 690 minat pH 11.4; this was approximately the upper limit ofpH for growth with doubling times below 12 h. Decreasinggrowth rates above pH 11 correlated with an apparent decrease in the ability to tightly regulate cytoplasmic pHand with the appearance of chains of cells. Whereas the cytoplasmic pH was maintained at pH 8.3 or below upto external pH values of 10.8, there was an increase up to pH 8.9 and 9.6 as the growth pH was increased to11.2 and 11.4, respectively. Both the transmembrane electrical potential and the phosphorylation potential(AGp) generally increased over the total pH range, except for a modest fall-off in the AGp at pH 11.4. Thecapacity for pH homeostasis rather than that for oxidative phosphorylation first appeared to become limitingfor growth at the high edge of the pH range. No cytoplasmic or membrane-associated organelles were observedat any growth pH, confirming earlier conclusions that structural sequestration of oxidative phosphorylationwas not used to resolve the discordance between the total electrochemical proton gradient (Ap) and the AGpas the external pH is raised. Were a strictly bulk chemiosmotic coupling mechanism to account for oxidativephosphorylation over the entire range, the AGp/Ap ratio (which would equal the H+/ATP ratio) would risefrom about 3 at pH 7.5 to 13 at pH 11.2, dropping to 7 at pH 11.4 only because of the rise in cytoplasmic pHrelative to other parameters. Moreover, the molar growth yields on malate were higher at pH 10.5 than at pH7.5, indicating greater rather than lesser efficiency in the use of substrate at the more alkaline pH.

Alkaliphilic Bacillus firmus OF4 grows at least as well at pH10.5 as at pH 7.5 in batch and continuous culture on malate-containing medium (5). At the more alkaline pH values, thecytoplasmic pH is maintained below pH 8.5, a remarkable,Na+-dependent capacity for pH homeostasis in which second-ary Na+/H+ antiporters apparently play a central role (13, 19).Having thus successfully confronted the primary biologicalchallenge, the alkaliphile must further resolve a bioenergeticproblem that results from this maintenance of a cytoplasmicpH (pHin) that is well below the external pH. Although asubstantial transmembrane electrical potential (Az4, positiveout) is generated by respiration-dependent proton extrusion,the magnitude of the electrochemical proton gradient (Ap,acid and positive out) is reduced by the chemiosmoticallyadverse pH gradient (ApH). Extreme alkaliphiles do not usethe much larger electrochemical gradient of sodium via asodium-coupled synthase, probably because outward sodiummovements must be coupled to proton accumulation forpurposes of pH homeostasis, and a primary sodium cyclewould subvert that major process (13). Since the F1Fo ATPsynthases of B. firmus OF4 (10) and at least one other extremealkaliphile (11) are thus exclusively proton coupled, there areonly two apparent ways to couple generation of a comparablephosphorylation potential (AGp) to bulk, delocalized chemi-osmotic driving forces that are about threefold greater at pH7.5 than at pH 10.5 (5). The oxidative phosphorylation machin-ery might be sequestered in organelles, but extensive investi-gation indicates no such arrangement (16) and, on the con-

* Corresponding author. Mailing address: Department of Biochem-istry, Box 1020, 1 G. Levy P1., New York, NY 10029. Phone: (212)241-7280. Fax: (212) 996-7214.

trary, points to a disperse localization of the ATP synthase ina conventional cytoplasmic membrane (26).

Alternatively, the alkaliphile synthase might use a variableH+/ATP stoichiometry. Since H+/ATP = AGp/Ap in a com-pletely chemiosmotically coupled scenario, an increase in thecoupling ratio would compensate for a decrease in the Ap athigher values of growth pH to allow generation of the sameAGp. Hoffmann and Dimroth (12) have indeed argued, on thebasis of measurements made on obligately alkaliphilic Bacillusalcalophilus at pH 10, that the H+/ATP ratio would not have tobe extremely high (only a little above 4) to account for theATP/ADP ratios that they found. However, as noted by us (6,13) and by Nicholls and Ferguson (22), studies at a single pHor even over a narrow range fail to assess just how large andvariable the H+/ATP stoichiometry would have to be in orderto account for the ATP synthesized. Also, the measurements ofthe ATP/ADP ratio and of the /\+ in the Hoffmann andDimroth study (12) were apparently not made under the sameconditions, which may have resulted in the recording ofATP/ADP ratios from cells in the original batch culture thathad lower AIj values than those aerated separately, in whichthe A/f measurements were made. Impressively, their data stillindicated that in B. alcalophilus, as in B. firmus OF4, thehighest ATP/ADP ratios were not observed at the highest Npvalues, as would be expected in a chemiosmotic couplingsituation.Both the Hoffmann and Dimroth studies on B. alcalophilus

(12) and earlier studies on B. firmus OF4 (5) were conductedin batch cultures, with no attempt made to rigorously establishan upper limit of pH for growth and thus ascertain how big thebioenergetic dilemma might be. In fact, no studies on alkali-philic Bacillus species in which growth was shown while themeasured pH was at 11 or more have been reported. Accord-

3111

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3112 STURR ET AL.

ingly, the factor(s) ultimately limiting growth is completelyunclear. Therefore, in this study, we sought to determine theupper limit for rapid growth of facultatively alkaliphilic B.firmus OF4 in continuous pH-controlled cultures and to study,for the first time, the bioenergetic parameters at various pHvalues in such cultures. By using malate-limited continuouscultures, the molar growth yields at pH 7.5 and pH 10.5 were

also to be determined, to see whether differences would be ina direction supporting a mechanism for oxidative phosphory-lation employing a variable H+/ATP, i.e., greater yields atlower pH. It was anticipated that these studies would furtherclarify the magnitude of the bioenergetic challenge to extremealkaliphiles with respect to oxidative phosphorylation andwould also indicate whether it was this capacity, problems ofpH homeostasis, or some other factor that limits growth at theupper edge of the pH range.

MATERIALS AND METHODS

Growth conditions. B. firmus OF4, isolated in our laboratory(4), was used for all studies. Growth in batch culture, forinocula and characterization of variants, was in the highlybuffered malate-containing medium previously described (5) atthe pH values indicated at 30°C.For continuous culture experiments, B. firmus OF4 was

grown at 30°C with vigorous aeration in a New BrunswickBioflow fermentor with a working volume of 350 ml. The same

carbonate-buffered medium used in the batch cultures was

used for cultures grown at highly alkaline levels, whereas a

phosphate-buffered medium that was otherwise of the same

composition was used for cultures at pH 8.5 or lower. Thecontinuous culture system was operated at constant pH by theadjustment of the medium in the reservoir to the target pHvalue for a given run and by the addition of NaOH or HCI as

needed during growth at particular pH values. An immersedpH electrode which acted as sensor for base or acid additionwas linked to a pump that delivered the required amount toreadjust the pH to the target level upon detection of deviation.The values detected by the immersed electrode were checkedfor accuracy against a calibrated external electrode whoseaccuracy had been verified against prepared buffers. Thecultures were maintained at steady state atA6. in the range of0.2 to 0.4. Continuous culture runs were conducted in at leastduplicate for every condition reported, and multiple samplingswere taken at each steady state.

Generation times. Generation times (tg) were calculatedfrom the flow rates that allowed the turbidity to remain withinthe 0.2 to 0.4 target range under any given condition. This wasdetermined empirically by adjusting the rate of addition offresh medium and monitoring A600. The most rapid rate ofmedium addition into the culture vessel that allowed mainte-nance of a steady A600 in the target range was determined over

long time frames (>4 h). Increasing the flow rate by 10% fromsuch values typically resulted in washout. The flow rates were

converted to dilution rates by dividing by the fixed volume ofthe fermentor, and the dilution rates were used as an index ofgrowth rate (,u) (2). The tg was calculated from the relationshiptg = ln 2/,u (2).

Bioenergetic measurements. Samples for determination ofAp and AGp were withdrawn from steady-state cultures of B.firmus OF4 growing continuously at the indicated pH value.Measurements of all parameters were conducted on the same

samples, in the growth medium, at the same cell concentrationsand assay conditions. ATP and ADP concentrations were

measured by luminometry as described previously (6), andphosphate concentrations were determined by the colorimetric

method of Lebel et al. (20). The values for AGp werecalculated by using values of AG' reported for the measuredcytoplasmic pH (27). The ApH component of the Ap wasdetermined from the distribution of weak acids or weak basesas described by others (32), and the transmembrane electricalpotential was similarly determined as recently described indetail (19) from the distribution of 3H-tetraphenylphospho-nium. Motility was assessed by phase-contrast microscopy.

Electron microscopy. For transmission electron microscopy,glutaraldehyde-fixed cells were postfixed in 1% osmium tetrox-ide, dehydrated, and embedded in epoxy resin (Epon; ElectronMicroscopy Sciences, Fort Washington, Pa.). Thick sectionswere stained with methylene blue and azure II. Smaller areaswere then selected for ultrastructural study. Thin sections werestained with uranyl acetate-lead citrate. They were examined ina JEM 100cx electron microscope (JEOL Ltd., Tokyo, Japan).The methods were as described by Hayat (8).Molar growth yields. Molar growth yields on D,L-malate-

containing media at either pH 7.5 or 10.5 were determined byusing continuous cultures at each of the two pH values thatcontained 0, 2.5, and 5.0 mM D,L-malate in the reservoirinstead of the standard concentration of 50 mM D,L-malate, butotherwise with the same medium as in the other experiments;the cultures grown in the absence of malate exhibited somegrowth that was attributable to the presence of 0.1% yeastextract in the complete medium. Malate consumption wasdetermined from measurements of both D-malate and L-malateconcentrations in culture samples from which cells had beenremoved by centrifugation. Total consumption was calculatedby subtracting the malate concentration in the medium in theculture vessel from that measured in the reservoir. The isomersof malate were assayed with stereospecific malate dehydroge-nases (Boehringer Mannheim) by monitoring NADH con-sumption spectrophotometrically.

For determinations of the dry weight, a calibration curverelating A600 to dry weight was established by weighing taredsamples of cultures that were prepared in the same medium atvarious A6. values. Cell samples were washed with distilledwater and then dried to constant weight in an oven. Dryweights of cells growing in continuous cultures were calculatedfrom measured A600 values, using the calibration curve. Allgrowth yield measurements were performed in triplicate onindependent experiments. The values for malate consumptionand biomass production obtained for growth on 0 mM malatewere used to correct the values at 2.5 mM malate. In eachexperiment, the value obtained for the 2.5 and 5.0 mM malatepair was the same as that calculated from the 0 and 2.5 mMmalate pair.

RESULTS

Growth and bioenergetic properties at various pH values.At pH 7.5, B. firmus OF4 grew with a tg of 54 min; the tgdecreased to 37 min at pH values from 8.5 to 10.6 (Fig. 1). ThepHi0 was 7.5 to 7.7 at values of external pH up to 9.5 and thenincreased to pH 8.2 or 8.3 at external pH values of 10.6 to 10.8.The tg increased to 94 min at pH 10.8, and upon a furtherincrease in the external pH to 11.2 and 11.4, marked increasesin tg, to 290 and 690 min, respectively, were observed. Theselatter increases in tg accompanied a significant rise in pHin, to8.9 and 9.6, the pattern suggesting a correlation between theincrease in tg and a decreasing ability to maintain a low pHin.Indeed, although data are not shown, variant cells were readilyisolated from the continuous culture at pH 11.2, with higherNa+/H+ antiport activity and a faster initial growth at pH 11than for the wild-type parent but no change in the cytochrome

J. BACTERIOL.

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

ALKALIPHILE GROWTH AND ENERGETICS 3113

AV,AP (mV)

-200-

-175-

-150-

-125-

-100-

-75-

-50-

-25-

0-

(i-)+1-

mod+++4 + +I

pHin

-10.0

AGp(mV)-600

--550

--500-450

-400

pHoutFIG. 1. Generation times, motility, and bioenergetic parameters of B. firmus OF4 growing in continuous cultures at different, constant values

of external pH. B. firmus OF4 was grown in continuous culture at various values of external pH (pH.u,) as described in Materials and Methods.The tg values and various bioenergetic parameters were measured as indicated in the text, and observations of motility were made (and arequalitatively reported at the top).

levels or A*i. As shown in Fig. 2, those pH values at which theslowest growth rates and the pronounced rise in pHin werefound also corresponded to conditions in which cells oftengrew in chains; nonetheless, at all pH values, a preponderanceof the cells were single or exhibited a single septum. Moreover,except for the chain formation, the fine structural featuresobservable in thin-section preparations were unchanged overthe very broad pH range studied, as shown for representativecells from cultures at pH 7.5 and 11.2 in Fig. 2. In particular, nointracellular or lens-like intramembrane structures (29) that

might sequester the oxidative phosphorylation machinery wereobserved.

B. firmus OF4 generated substantial A* values over theentire pH range, with a gradual increase from -140 mV at pH7.5 to -182 mV at pH 10.6, followed by a peak values of -200mV at pH 10.8 before a return to the range of -182 to -185mV at pH 11.2 and 11.4 (Fig. 1). The Ap declined significantlybetween values of the growth pH of 7.5 and 10.6, because ofthe increasing acidification of the cytoplasm relative to themedium pH. As found previously (e.g., reference 6) for cells

C

35m0.5 mm

B D

_irM.mrqw_ @p

_ __oi * d 10 UM _ - 0.5 m

FIG. 2. Thin sections of cells of B. firmus OF4 grown in continuous culture at pH 7.5 or 11.2. Representative fields of cells from cultures at thetwo pH values are shown in panels A and B, and a single, representative cell from each culture is shown at higher magnification in panels C andD. (A and C) pH 7.5; (B and D) pH 11.2.

VOL. 176, 1994

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3114 STURR ET AL.

04

0

.4

1 5

12.5

10-

7.5

5

2.5 47 8 9 10 11 12

pH out

FIG. 3. The AGp/Ap ratio of B. firmus OF4 as a function of theexternal pH of continuous cultures. The data are calculated from thosepresented in Fig. 1.

incubated in the pH range around 10.5, the observation of ahigher Aqp in a particular sample set, e.g., -200 mV at pH 10.8,was accompanied by a correspondingly lower pHi., presumablybecause the electrogenic antiporters that account for intracel-lular proton accumulation are energized by the ,A. As a result,the total Ap was steady at about -50 mV between pH 10.6 and11.2, increasing at pH 11.4 only because of the rise in pHin. Bycontrast, the AGp increased over almost the entire pH range,decreasing only between pH 11.2 and 11.4. The AGp/Ap ratiois shown as a function of growth pH in Fig. 3. At pH 7.5, theratio was approximately 3, rising precipitously until the highestvalue of 13 was reached at pH 11.2. The parameter thatchanged the most in these calculations was the external pH,which was set over a range of almost four full pH units,whereas the pHin varied in a range of only slightly over two pHunits, and changes in the AGp and At were comparativelymore modest. Also, interestingly, the AGp/Ap ratio and puta-tive H+/ATP ratio decreased to 7 at pH 11.4, partly because ofa reduction in the AGp but largely because of the increase inthe calculated Ap that resulted entirely from an apparentfailure of pH homeostasis.A second energetic process in B. firmus OF4 that is coupled

to the AiJ, but via Na+ rather than protons, is motility (3a, 17).As shown qualitatively at the top of Fig. 1, motility, as observedby phase-contrast microscopy, was greatest at pH 10.6 and10.8. No motility was observed at pH 7.5, and it is not knownwhether B. firmus OF4 synthesizes flagella at that pH.Molar growth yields on malate. The molar growth yield on

malate for B. firmus OF4 cultured continuously at pH 7.5 wasdetermined to be 53 mg (dry weight)/mmol of malate con-sumed (Table 1). At pH 10.5, a value of 82 mg (dry weight)/

TABLE 1. Molar growth yields of B. firinus OF4 on malateat pH 7.5 and 10.5

Molar growth yield' Efficiency' Preference ratio (mmol ofpHout (mg [dry wtJ/mmol of Ein L-malate consumed/mmol

malate consumed) of D-malate consumed)

7.5 53 43 5:110.5 82 66 3:1

aDetermined as described in Materials and Methods.b Determined as milligrams (dry weight) produced per milligram of malate

consumed.

mmol of malate was found. These values are higher thanreported previously for batch cultures, and the differencebetween pH 7.5- and pH 10.5-grown cells is greater (5). Inthose earlier studies, cells were growth limited by the malateconcentrations, but the actual malate consumption was notdetermined. In the current experiments, the higher efficiencyof malate utilization at pH 10.5, 66%, compared with 43% atpH 7.5, accompanied an enhanced ability to use D-malate atthe higher pH, as indicated by the lower preference ratio of 3:1for D-malate versus L-malate at the higher pH (Table 1).

DISCUSSION

Alkaliphilic B. firnus OF4 is clearly capable of robustnonfermentative growth at pH values above 11. Even theslowest rates of growth observed in this study were well withinthe range for many conventional neutralophiles. Moreover, itis likely that obligately alkaliphilic strains grow even betterthan facultatively alkaliphilic B. firnus OF4, since the mem-brane lipids of obligately B. firmus RAB, for example, areapparently better optimized for growth in the very alkalinerange of pH at the expense of a capacity for growth in the nearneutral range (3). At the upper edge of the pH range that wascompatible for rapid growth ofB. firnus OF4 in the currentstudy, the most likely limiting factor was the capacity for pHhomeostasis. In an average of several determinations, therewas a small initial increase in tg at pH 10.8 that precededobservations of markedly increased values of cytoplasmic pH.However, the conditions under which the tg values werecalculated may have been at slight variance from those atwhich the bioenergetic parameters were measured. The tg wascalculated directly from cells growing in the primary culturevessel, while it was necessary to remove samples from thisvessel in order to measure the bioenergetic parameters. Withthat small initial exception, the cytoplasmic pH was the param-eter whose change best correlated with increasing tg in the mostalkaline pH range. A physiological process that appeared to becompromised concomitantly was normal septation; formationof chains has similarly been observed during growth of anaer-obic, alkaliphilic Clostridium paradoxum at the upper end of itspH range (21). There may be some step in septation that isgenerally sensitive in bacilli as the cytoplasmic pH rises, or anSOS-like response that includes an effect on division takesplace at the alkaline edge of cytoplasmic pH, as has beenreported for Escherichia coli (28).Above pH 11.0, B. firnus OF4 exhibited no decline in the A*J

relative to that at lower pH values. The capacity for oxidativephosphorylation, as indicated by the magnitude of the AGp,was even higher at pH 11.2 than at lower pH values, falling offslightly only after the external pH was raised further so that pHhomeostasis was compromised markedly. Thus, it is unlikelythat either the primary generation of a Ap or the synthesis ofATP was the major limiting factor in the rate of growth at theupper edge of the pH range. Oxidative phosphorylation by theextreme alkaliphile was remarkably resistant to a very alkalinemedium pH and to cytoplasmic pH values above 9; thisobservation was consistent with earlier findings in experimentswith starved cells that were poised at different pH values (6).That the capacity for pH homeostasis was indeed the limitingfactor above pH 11 was supported by the finding of variantswith increased Na+/H+ antiport activity.There was a large change in the AGp/Ap over the range of

growth pH of B. firmus OF4. The values found at pH 7.5 and10.5 correlated very well with the values reported earlier forbatch cultures (5), but the broader range of pH used hererevealed a much greater variation in the ratio with pH than

J. BAcrERIOL.

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

ALKALIPHILE GROWTH AND ENERGETICS 3115

found before. Moreover, it was apparent under these carefullycontrolled growth conditions that the parameter that contrib-uted most substantially to the change in AGp/Ap ratio was theexternal pH. This parameter was set, controlled, and moni-tored with both the internal electrode and regular periodicmeasurements with a calibrated external electrode during eachexperiment. Numerous controls and standardizations indicateno pH-dependent loss of accuracy in these measurements, norin the less direct determinations of the bioenergetic parame-ters that exhibited smaller fluctuations (13, 19).Again as in earlier studies (16, 26), there was no indication

of cytoplasmic or membrane-associated organelles that couldsequester the machinery for oxidative phosphorylation fromthe transcytoplasmic membrane gradients that diminish withincreasing growth pH. If a purely chemiosmotic mechanism,with complete coupling of oxidative phosphorylation to thebulk delocalized Ap, obtained over the whole pH range, theH+/ATP ratio would therefore rise from about 3 at pH 7.5 upto 13 at pH 11.2. The H+/ATP coupling ratio would thenputatively fall to 7 at pH 11.4. That unlikely conclusion derivesfrom a calculation that was affected primarily by the failure ofpH homeostasis rather than anything directly related to oxida-tive phosphorylation. Marked differences in the H+/ATP sto-ichiometry of a given H+-coupled synthase have been reportedin reconstituted systems in which the lipid was varied (31), andKrenn et al. (18) have recently suggested a correlation betweenthe H+/ATP stoichiometry of a given ATP synthase and thecontent and pattern of charged residues in the primary struc-tures of Fo subunits c and a. The current experiments do notexclude the use of an enormously variable H+/ATP ratio in thealkaliphile. However, several points should be noted. First,there are no significant differences between the membranelipid composition of pH 7.5- and pH 10.5-grown cells of B.firnus OF4 (1). Second, both molecular biological and bio-chemical data indicate that B. firmus OF4 possesses a singleATP synthase species, which is the product of a unique atpoperon and is assembled in an invariant subunit stoichiometryand membrane concentration (12a, 14). Finally, were theH+/ATP ratio to increase from 3 to 13, there would have to besome combination of an increasing H+/O and a reduced molargrowth yield to accommodate the changing H+/ATP ratio. Infact, there are high pH-dependent increases in several respi-ratory chain components (24), with the increase in the caa3-type terminal oxidase having been shown to occur at thetranscriptional level (25). However, there is no evidence for aprogressive increase in the possible proton-translocating com-plexes over a broad range of pH values. Thus, to accommodatean increasing H+/ATP, at least some diminution in the molargrowth yield would be expected. Just the opposite was ob-served, with a substantially greater molar growth yield onmalate found at pH 10.5 than at pH 7.5. The molar growthyields at both pH values, but especially at pH 10.5, were in thehigh end of the range of values reported for aerobic growth ofbacteria on Krebs cycle intermediates (23), e.g., 61.8 mg (dryweight)/mmol of citrate consumed for Aerobacter aerogenes (7)and 34.8 mg (dry weight)/mmol of malate consumed forPseudomonas fluorescens (9). Other models for the energiza-tion of oxidative phosphorylation in the alkaliphile at pHvalues between 10 and 11.4 must in any event take account ofseveral qualitative observations that are also difficult to recon-cile with a strictly chemiosmotic model. Alternative models,including delocalized proton movements on the membranesurface, e.g., as proposed by others (15, 30), or direct protontransfers between a respiratory chain complex and the ATPsynthase (6), are under examination.

ACKNOWLEDGMENTSWe acknowledge and thank Norman Katz for the electron micros-

copy work and Jeanne Poindexter for providing part of the continuousculture equipment and many helpful comments related to its use.

This work was supported by research grant GM28454 from theNational Institute of General Medical Sciences.

REFERENCES1. Clejan, S., T. A. Krulwich, K R. Mondrus, and D. Seto-Young.

1986. Membrane lipid composition of obligately and facultativelyalkalophilic strains of Bacillus. J. Bacteriol. 168:334-340.

2. Drew, S. W. 1981. Liquid culture, p. 151-178. In P. Gerhardt,R. G. E. Murray, R. N. Costilow, E. W. Nester, W. A. Wood, N. R.Krieg, and G. B. Phillips (ed.), Manual of methods for generalbacteriology. American Society for Microbiology, Washington,D.C.

3. Dunkley, E. A., Jr., A. A. Guffanti, S. Clejan, and T. A. Krulwich.1991. Facultative alkaliphiles lack fatty acid desaturase activity andlose the ability to grow at near neutral pH when supplementedwith an unsaturated fatty acid. J. Bacteriol. 173:1331-1334.

3a.Guffanti, A. Unpublished data.4. Guffanti, A. A., 0. Finkelthal, D. B. Hicks, L. Falk, A. Sidhu, A.

Garro, and T. A. Krulwich. 1986. Isolation and characterization ofnew facultatively alkalophilic strains of Bacillus. J. Bacteriol.167:766-773.

5. Guffanti, A. A., and D. B. Hicks. 1991. Molar growth yields andbioenergetic parameters of extremely alkaliphilic Bacillus speciesin batch cultures, and growth in a chemostat at pH 10.5. J. Gen.Microbiol. 137:2375-2379.

6. Guffanti, A. A., and T. A. Krulwich. 1992. Features of apparentnon-chemiosmotic energization of oxidative phosphorylation byalkaliphilic Bacillus firmus 0F4. J. Biol. Chem. 267:9580-9588.

7. Hadjipetrou, L. P., J. P. Gerrits, F. A. G. Teulings, and A. H.Stouthamer. 1964. Relation between energy production andgrowth of Aerobacter aerogenes. J. Gen. Microbiol. 36:139-150.

8. Hayat, M. A. 1989. Principles and techniques of electron micros-copy. CRC Press, Inc., Boca Raton, Fla.

9. Hernandez, E., and M. J. Johnson. 1967. Energy supply and cellyield in aerobically grown microorganisms. J. Bacteriol. 94:996-1001.

10. Hicks, D. B., and T. A. Krulwich. 1990. Purification and reconsti-tution of the FIFO-ATP synthase from alkaliphilic Bacillus firmusOF4: evidence that the enzyme translocates H+ but not Na+. J.Biol. Chem. 265:20547-29554.

11. Hofnann, A., and P. Dimroth. 1991. The ATPase of Bacillusalcalophilus. Reconstitution of energy-transducing functions. Eur.J. Biochem. 196:493-497.

12. Hofmnann, A., and P. Dimroth. 1991. The electrochemical protonpotential of Bacillus alcalophilus. Eur. J. Biochem. 201:467-473.

12a.Ivey, D. M. Unpublished data.13. Ivey, D. M., A. A. Gufanti, and T. A. Krulwich. 1992. The Na+

cycle of alkaliphilic Bacillus species, p. 101-124. In E. P. Bakker(ed.), Alkali, cation transport systems. CRC Press, Inc., BocaRaton, Fla.

14. Ivey, D. M., and T. A. Krulwich. 1992. Two unrelated alkaliphilicBacillus species possess identical deviations in sequence fromthose of conventional prokaryotes in regions of Fo genes impli-cated in proton translocation through the ATP synthase. Res.Microbiol. 143:467-470.

15. Kates, M., J.-Y. Syz, D. Gosser, and T. H. Haines. 1993. pH-dissociation characteristics of cardiolipin and its 2'-deoxy ana-logue. Lipids 28:877-882.

16. Khan, S., D. M. Ivey, and T. A. Krulwich. 1992. Membraneultrastructure of alkaliphilic Bacillus species studied by rapidfreeze electron microscopy. J. Bacteriol. 174:5123-5126.

17. Kitada, M., A. A. Gufanti, and T. A. Krulwich. 1982. Bioenergeticproperties and viability of the alkalophilic Bacillus firmus RAB asa function of pH and Na+ content of the medium. J. Bacteriol.152:1096-1104.

18. Krenn, B. E., H. S. Van Wairaven, M. J. C. Scholts, and RKraayenhof. 1993. Modulation of the proton-translocation stoichi-ometry of H+-synthases in two phototrophic prokaryotes byexternal pH. Biochem. J. 294:705-709.

VOL. 176, 1994

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3116 STURR ET AL.

19. Krulwich, T. A., and A. A. Guffanti. 1992. Proton-coupled bioen-ergetic processes in extremely alkaliphilic bacteria. J. Bioenerg.Biomembr. 24:587-599.

20. Lebel, D., G. G. Poirer, and A. R Beaudoin. 1978. A convenientmethod for the ATPase assay. Anal. Biochem. 85:86-89.

21. Li, Y., L. Mandelco, and J. Wiegel. 1993. Isolation and character-ization of a moderately thermophilic anaerobic alkaliphile, Clos-tridium paradoxum sp. nov. Int. J. Syst. Bacteriol. 43:450-460.

22. Nicholls, D. G., and S. J. Ferguson. 1992. The chemiosmoticcircuit, p. 97-98. In Bioenergetics 2. Academic Press, London.

23. Payne, W. J. 1970. Energy yields and growth of heterotrophs.Annu. Rev. Microbiol. 24:17-52.

24. Quirk, P. G., A. A. Guffanti, R J. Plass, S. Clejan, and T. A.Krulwich. 1991. Protonophore-resistance and cytochrome expres-sion in mutant strains of the facultative alkaliphile Bacillus firmusOF4. Biochim. Biophys. Acta 1058:131-140.

25. Quirk, P. G., D. B. Hicks, and T. A. Krulwich. 1993. Cloning of thecta operon from alkaliphilic Bacillus firmus OF4 and characteriza-tion of the pH-regulated cytochrome caa3 oxidase it encodes. J.Biol. Chem. 268:678-685.

26. Rohde, M., F. Mayer, D. B. Hicks, and T. A. Krulwich. 1989.Immunoelectron microscopic localization of the FIFO ATPase(ATP synthase) on the cytoplasmic membrane of alkalophilic

Bacillus firmus RAB. Biochim. Biophys. Acta 985:233-235.27. Rosing, J., and E. C. Slater. 1972. The value of AG for the

hydrolysis of ATP. Biochim. Biophys. Acta 267:275-290.28. Schuldiner, S., V. Agmon, J. Brandsma, A. Cohen, E. Friedman,

and E. Padan. 1986. Induction of SOS functions by alkalineintracellular pH in Escherichia coli. J. Bacteriol. 168:936-939.

29. Skulachev, V. P. 1991. Chemiosmotic systems in bioenergetics: H+cycles and Na+ cycles, p. 387-444. In P. Mitchell and C. A.Pasternak, (ed.), Perspectives in vectorial metabolism and osmo-chemistry. The Glynn Research Foundation, Glynn, United King-dom.

30. Tocanne, J. F., and J. Teissie. 1990. Ionization of phospholipidsand phospholipid-supported interfacial lateral diffusion of protonsin membrane model systems. Biochim. Biophys. Acta 1031:111-142.

31. van Walraven, H. S., M. H. C. Scholts, F. Koppenaal, R H. A.Bakels, and K. Krab. 1990. Dependence of the proton transloca-tion stoichiometry of cyanobacterial and chloroplast H+-ATPsynthase on the membrane composition. Biochim. Biophys. Acta1015:425-434.

32. Zilberstein, D., V. Agmon, S. Schuldiner, and E. Padan. 1982. Thesodium/proton antiporter is part of the pH homeostasis mecha-nism in Escherichia coli. J. Biol. Chem. 257:3687-3691.

J. BACrERIOL.

on March 9, 2021 by guest

http://jb.asm.org/

Dow

nloaded from